| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Neuroradiology |
1 From the Second Departments of Diagnostic Radiology (C.J.C, H.L.H, Y.C.T., Y.C.W., L.J.W.), Orthopedics (C.C.N.), and Neurosurgery (T.Y.C.), Chang Gung Memorial Hospital, 199 Tung-Hwa North Rd, Taipei, Taiwan; and Public Health and Biostatistics Center, Chang Gung University, Taoyuan, Taiwan (M.C.C.). Received February 18, 2002; revision requested May 15; final revision received July 12; accepted August 8. C.J.C. supported by Taiwan National Science Council grant 89-2314-B-182A-086 and Chang Gung Memorial Hospital grant CMRP 968. Address correspondence to (e-mail: radcjc@adm.cgmh.org.tw).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Sixty-two patients with cervical degenerative disease were evaluated with regard to the dynamic changes of canal stenosis at flexion-extension MR imaging. Functional cord impingement was considered if the cord was impinged or more impinged after neck flexion or extension. Selection criteria for neutral-position MR imaging, such as cervical curvature, canal space, degenerative stage, intramedullary high signal intensity on T2-weighted images, and resting instability, were evaluated for their ability to predict functional cord impingement at flexion-extension MR imaging (Fisher exact test, logistic regression analysis).
RESULTS: MR images in 19 (31%) of 62 patients showed functional cord impingement at extension MR imaging compared with images in two (3%) patients at flexion MR imaging. Statistically significant differences were found for the criteria cervical degeneration stage (P < .001) and spinal canal space (P = .037) for predicting functional cord impingement at extension MR imaging. In contrast, no significant differences were found among selection criteria for flexion MR imaging. Probabilities of functional cord impingement at extension MR imaging were calculated with different combinations of degenerative stages and canal spaces. Probability could increase to 79% if the patient had both stabilization degeneration (disk protrusion or osteophytic formation with hypertrophy of the ligamentum flavum) and C7 canal space of 10 mm or less.
CONCLUSION: None of the selection criteria evaluated in this study has the ability to predict functional cord impingement at flexion MR imaging; however, prediction of impingement at extension MR imaging can increase from 31% to 79% if proper criteria are selected.
© RSNA, 2003
Index terms: Spine, MR, 31.121412
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Patient Positioning
All patients were examined with a 1.5-T superconducting MR unit (Signa; GE Medical Systems, Milwaukee, Wis). After a routine supine neutral-position (0°) examination, the dynamic examination was performed with neck flexion (40°) and extension (-20°) by using custom-built positioning sponges. A posterior neck receive-only surface coil was placed beneath the sponges. Sometimes with flexion and extension, the patient would move his or her head into an off-sagittal position. This problem was corrected by means of repeated adjustments until the radiologist was satisfied.
MR Imaging Protocol
The routine supine (0°) MR imaging protocol consisted of sagittal T1-weighted (spin-echo: 466/11 [repetition time msec/echo time msec]), sagittal T2-weighted (fast spin-echo: 4,000–4,133/80–85), sagittal T2*-weighted (gradient-echo: 483/17, flip angle 20°), and transverse T2*-weighted (gradient-echo: 433/17, flip angle 20°) MR sequences. The flexion-extension MR imaging protocol consisted of sagittal T2-weighted (fast spin-echo) and transverse T2*-weighted (gradient-echo) sequences in both flexion (40°) and extension (-20°) postures. Matrix size was 256 x 192 for transverse images and 256 x 256 for sagittal images. Two signals were acquired. Section thickness was 4 mm with 1-mm gap for both sagittal and transverse MR images.
Evaluation
The evaluation included two parts. Part one was assessment of the dynamic change in spinal canal stenosis during neck flexion and extension. Part two was definition of imaging criteria that might predict functional cord impingement at flexion-extension MR imaging.
To ascertain the dynamic changes, the extent of spinal stenosis in neutral, flexed, and extended postures was graded as follows: grade 0, no obliteration or partial obliteration of the anterior or posterior subarachnoid spaces; grade 1, complete obliteration of the anterior or posterior subarachnoid spaces; and grade 2, anterior and/or posterior cord impingement (pincer effect). All cervical segments were evaluated, and the most stenotic segment was recorded. Functional cord impingement was considered if the cord was impinged or was more impinged after neck flexion or extension. "More impinged" was defined by means of measurement of a reduction in the cord size after neck flexion or extension, which was considered a positive finding. This evaluation was performed by two radiologists (L.J.W., Y.C.W.) independently without knowledge of the patient’s history and clinical findings. Disagreements were recorded and resolved later with consensus.
In regard to neutral-position MR imaging criteria that will predict functional cord impingement at flexion-extension MR imaging, the following criteria were evaluated: types of cervical curvature, stages of cervical degeneration, signs of resting instability and space of the C7 spinal canal, and the existence of intramedullary high signal intensity on T2-weighted MR images. Two radiologists (C.J.C., H.L.H.), who were blinded to the clinical status, were asked to classify these imaging criteria together. Disagreements were resolved with consensus.
Cervical curvature was classified according to the principles suggested by Guigui et al (6) and Batzdorf and Batzdorff (7). Cervical curvature was measured according to the relation of the dorsal aspect of C3 through C6 to a line drawn from the dorsocaudal aspect of C2 to that of C7. By definition, type 0 is a normal lordotic cervical curvature in which no part of the dorsal aspect of the C3 through C6 crosses the C2 through C7 line. Type 1 is a straight or meandering cervical curvature in which all of the dorsal aspects of C3 through C6 meet with the C2 through C7 line or partially cross it. Type 2 is a kyphotic curvature in which all the dorsal aspects of C3 through C6 cross the C2 through C7 line.
Cervical degeneration was classified into the following stages (8): stage 1, discogenic phase, characterized at MR imaging as normal, small annular tears or disk protrusion without osteophytic formation; stage 2, spondylosis, characterized at MR imaging on the basis of disk protrusion with osteophytic formation; and stage 3, stabilization phase, characterized at MR imaging as disk protrusion or osteophytic formation with hypertrophy of the ligamentum flavum. In the stabilization phase, the posterior element involvement, including joint arthrosis, laminar hypertrophy, and ligamentum flavum hypertrophy, plays a more obvious role. In this study, ligamentum flavum hypertrophy was considered a sign of posterior element involvement because it was detected more easily than were joint arthrosis or laminar hypertrophy at MR imaging. The problem of differentiating stage 1 from stage 2 degeneration, which appears to relate solely to the absence or presence of osteophyte formation, respectively, was aided by means of T2*-weighted MR imaging because of its better resolution of osteophytes. All the cervical segments were evaluated, and the segment with the most advanced degenerative stage was recorded.
Spinal canal space was defined as the diameter of C7 at the midvertebral level of the most midsagittal section. C7 was chosen because it was less involved with spondylosis and might represent the true canal diameter of the middle and lower cervical spine. The C7 canal space was classified into three categories: normal, sagittal diameter of 13 mm or more; mild narrowing, sagittal diameter of 11–12 mm; and moderate to severe narrowing, sagittal diameter of 10 mm or less.
Potential instability can be observed in resting (neutral) posture, depending on two factors: range of angular intervertebral mobility and horizontal displacement of the vertebral body (6,9). In neutral position, the degree of horizontal displacement of the vertebral body is expressed as the distance between two lines that extend from the posterior margin of the lower vertebrae and the posteroinferior margin of the displaced vertebra (9,10). The range of angular intervertebral mobility is expressed as the angle between two lines that extend from the inferior margins of two adjacent vertebrae (6,9). In neutral position, the existence of potential instability (resting instability) is considered if horizontal displacement of 2.7 mm or more or a range of angular intervertebral mobility of 11° or more is observed in any segment of the cervical spine (9,10).
The existence of intramedullary high signal intensity on T2-weighted images is considered if any abnormal high signal intensity is noted in the cervical cord.
Statistical Analysis
Data were analyzed with a software program (SAS, version 6.12; SAS Institute, Durham, NC). Interobserver agreement for the grading of spinal stenosis was evaluated and expressed with 
statistics. Agreement was said to be excellent with > 0.80, good with = 0.80–0.61, moderate with = 0.60–0.41, fair with = 0.40–0.21, and poor with < 0.20. The 
2 test or Fisher exact test was used to examine the relationship between patients with functional cord impingement and the imaging criteria. Differences with P < .05 were considered statistically significant.
Logistic regression analysis was used to evaluate the relationship between functional cord impingement and a set of explanatory variables, including possible confounding effects such as patient age and sex and the factors of interest, including spinal canal space, cervical degeneration, cervical curvature, instability, and intramedullary high signal intensity on T2-weighted images. For each variable, male sex, lordotic curvature, discogenic degeneration, spinal canal space of 13 mm or more, no sign of instability, and no intramedullary high signal intensity on T2-weighted MR images were treated as baseline. After the final logistic model was performed, probability was calculated of functional cord impingement at extension or flexion MR imaging with different combinations of variables.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
= 0.787).
Neutral-Position MR Imaging Criteria to Predict Functional Cord Impingement
Nineteen (31%) of 62 patients had functional cord impingement at extension MR imaging compared with two (3%) at flexion MR imaging. The Fisher exact test showed a significant difference (P < .001) between flexion and extension MR imaging with regard to functional cord impingement.
With regard to cervical curvature, five (23%) of 22 patients with lordotic curvature, 10 (36%) of 28 patients with straight or meandering curvature, and four (33%) of 12 patients with kyphotic curvature had functional cord impingement at extension MR imaging, while one (5%), one (4%), and none, respectively, had functional cord impingement at flexion MR imaging. The Fisher exact test showed no significant difference between these three group with regard to functional cord impingement at extension (P = .631) and flexion (P = 0.99) MR imaging (Table 1).
|
With regard to cervical degeneration, one (5%) of 21 patients with discogenic degeneration, three (21%) of 14 patients with spondylosis degeneration (Fig 2), and 15 (56%) of 27 patients with stabilization degeneration (Fig 3) had functional cord impingement at extension MR imaging. A significant difference was found among these three groups with regard to functional cord impingement (P < .001) at extension MR imaging. None of the patients with discogenic or spondylosis degeneration and two (7%) of the 27 patients with stabilization degeneration had functional cord impingement at flexion MR imaging. No significant difference was found between these three groups with regard to functional cord impingement (P = .5) at flexion MR imaging (Table 1).
|
|
With regard to intramedullary high signal intensity on T2-weighted MR images, 17 (29%) of 59 patients without and two (67%) of three patients with the finding had functional cord impingement at extension MR imaging. One (2%) of 59 patients without and one (33%) of three patients with the finding had functional cord impingement at flexion MR imaging. No significant difference was found between these groups with regard to functional cord impingement at extension (P = .220) and flexion (P = .095) MR imaging (Table 1).
After adjustment for the confounding effect from other variables by means of logistic regression, a significant difference was found between cervical degenerative stages (P = .0364) and the trend for significant differences (P = .0820) between canal spaces with regard to functional cord impingement at extension MR imaging (Table 2). In contrast, no significant factor was found at flexion MR imaging. To simplify the model and calculate probability of functional cord impingement at extension MR imaging, most of the factors that were not significant, such as age, sex, cervical curvature, instability, and intramedullary high signal intensity on T2-weighted images were deleted from the final logistic regression model. Table 3 shows that for functional cord impingement at extension MR imaging, odds ratios of canal space of 13 mm or more to canal spaces of 11–12 mm and 10 mm or less are 2.5 and 8.2, respectively. Odds ratios of discogenic degeneration to spondylosis and stabilization for functional cord impingement at extension MR imaging are 7.3 and 25.5, respectively.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Data from the 62 patients in the present study and from the 81 patients in the study by Muhle et al (16) show that increased cervical cord impingement was seen in 27% (22 of 81) to 31% (19 of 62) of extension MR images and in 3% (two of 62) to 5% (four of 81) of flexion MR images. Selection criteria are needed to improve the effectiveness of flexion-extension MR imaging.
With regard to flexion-extension cervical MR imaging, one may suspect that the degrees of flexion and extension are limited inside the bore of an MR imager compared with those allowed at flexion-extension radiography. After comparing active and passive segmental movement of the cervical spine at flexion-extension radiography and flexion-extension MR imaging, Muhle et al (14) reported only small differences between them. The low coefficient of variance between repeated examinations in their study also indicates that flexion-extension MR imaging is a tool with good reproducibility. The main reason to acquire a flexion-extension radiograph is to evaluate for instability. Flexion-extension MR imaging provides information about instability and dynamic changes in hypotonic ligamentum flavum and annular bulging and, most important, the effect of these changes in the spinal cord. In this study, we focused on evaluating neutral-position criteria that may help improve the effectiveness of flexion-extension MR imaging. Further studies are needed to investigate the clinical value of this technique.
At flexion cervical MR imaging, 3% (two of 62) of patients had functional cord impingement. More patients (14% [nine of 62]) had a decrease in spinal stenosis, even with cord decompression, at flexion MR imaging. These results may indicate a minor role of flexion in contributing to functional cord impingement. The same observation was also reported by other investigators (12,16). After we evaluated these possible selection criteria, we found that none had the ability to improve the effectiveness of flexion MR imaging.
At extension MR imaging, cervical degenerative stage was found to play a major role. Cervical degeneration is a continual process with four phases: dysfunction, disk degeneration, spondylosis, and stabilization (8,17). Because of the relatively preserved integrity of the cervical spine in the discogenic phase, 5% (one of 21) of patients had functional cord impingement at extension MR imaging. As the loss of disk height and osteophytic formation proceed in the spondylosis phase, segmental instability may occur during cervical motion. An increase (to 21% [three of 14]) in the number of cases of functional cord impingement at extension MR imaging was found in patients in our study with spondylosis degeneration. When disease progresses to the stabilization phase, posterior element involvement plays a more obvious role. In the beginning of this phase, progression of osteophytosis can lead to restricted motion of the involved cervical segments with secondary instability of adjacent cervical levels. To increase joint stability, secondary posterior element changes, such as facet joint arthrosis and laminar and ligamentum flavum hypertrophy, may occur. Since these structures are positioned centrally, the posterior space available for the cervical cord is decreased. This is why we see a dramatic increase in the incidence of functional cord impingement at extension MR imaging. The odds ratio of the discogenic phase to the stabilization phase for functional cord impingement at extension MR imaging was high, up to 25.5.
At univariate analysis, spinal canal space was also found to play a role, though smaller, in predicting the value of extension MR imaging (P = .037). It is well known that the size of the spinal canal is an important etiologic factor for cervical spondylotic myelopathy (18). It is also reported that development of cord compression as a result of spondylotic change is unlikely when the sagittal diameter of the canal is larger than 13 mm (19). We also saw this trend in the present study. The odds ratio of C7 canal space of 13 mm or more to C7 canal space of 10 mm or less for functional cord impingement at extension MR imaging was 8.2.
Statistically, we found that the variables of cervical curvature, signs of resting instability, and intramedullary high signal intensity on T2-weighted images played no role in predicting the effectiveness of extension MR imaging. Though 67% (two of three) of patients with intramedullary high signal intensity had functional cord impingement at extension MR imaging, the insignificant variables related to only these three patients.
Probabilities of functional cord impingement at extension MR imaging with different combinations of C7 canal spaces and degenerative stages were widely variable because they were predictions from a relatively small data set. However, by combining results of univariate and multivariate analyses, a simple guideline can be drawn from these data. An additional extension MR imaging examination is suggested in patients with signs of ligamentum flavum hypertrophy (stabilization phase) or of both spondylosis degeneration and C7 canal space of 10 mm or less because the probability of functional cord impingement is greater than average (47% [29 of 62] of patients in this group). In patients with probabilities smaller than average (53% [33 of 62]), extension MR imaging can be deleted initially and performed later if the clinicians suggest, after correlation with clinical findings, that such an examination is needed. This guideline is simple and easily applied. Radiologists can easily decide whether to perform extension MR imaging before the final pulse sequence in a neutral-position examination.
In conclusion, none of the selection criteria used in this study has the ability to predict functional cord impingement at flexion MR imaging; however, prediction of impingement at extension MR imaging can increase from 31% to 79% if proper criteria are selected.
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
